Hello !

Feb 2016 - Present (Tempe, United States)
Warehouse as a Service: The product helps sellers manage their entire warehouse inventory and ship products from their own warehouse.
Tech Stack: Java, ReactJS, MobX, DynamoDB, Elastic Search, Mockito, JEST, Amazon build and deployment tools

July 2015 - June 2016 (Pune, India)
Krixi Care Product: Enables physicians to remotely evaluate, diagnose and treat acute medical conditions. (Duration: 7 months)
Led the team of 5 developers and assigned tasks. Implemented RESTful APIs for scheduling consult with patients.
Integrated Stripe payment gateway & real-time calendar update feature. Responsible for deployments.
Technologies- Grails, Groovy, Tomcat, MySQL, MongoDB

August 2014 - June 2016 (Pune, India)
EMR Architecture Project: Developed a scalable lifecycle POC to process EMRs from healthcare practices in US (Duration:12 months)
Increased performance by 400% using Spark as compared to earlier model.
Improvised resource allocation by introducing YARN. Managed distributed infrastructure on the servers.
Technologies: Java, Apache Spark, Hadoop, Solr, Cassandra, Maven

Citrus Wallet: Web app for consumers to load money from cards or net banking accounts to speed up payments. (Duration: 3 months)
Developed Single Page app for a CitrusPay payment gateway client to provide e-wallet features for their consumers.
Integrated key features– load/withdraw money, profile, OTP, transaction history.
Technologies: Groovy, Grails, MySQL, Redis, JQuery, Javascript

Oct 2013 - July 2014 (Pune, India)
Vodafone-Essar Limited BI: Manage the network data of Vodafone India on client location IBM. Part of BI-3G team.
Resolved issues in the ETL process and maintained its data warehouse. Technologies- Oracle, Teradata, Datastage, shell scripts

August 2016 - November 2018. EMITLab Research Group (Tempe, US)
MS Thesis - Load-balanced Range Query Workload Partitioning for Compressed Spatial Hierarchical Bitmap (cSHB) Indexes
(Defended Nov 2018; Committee: Dr. Selcuk Candan and others)
Proposed novel partitioning techniques to efficiently partition range query workload into balanced partitions. The partitions can be processed in parallel using spatial bitmap indexes that leverage block-based hierarchy in bitmap index. Introduced a cost model to quickly group the queries based on their distance and coverage using R*-Tree spatial data structure.
Technologies- Java, Matlab

DataStorm: (NSF Project) End-to-end disaster planning and response. Explored methods to process workflows in parallel and record provenance data for simulations in disaster models-Flood, Hurricane. Worked on aggregated dynamic visualization of all models. Mentored undergraduate students for creating layered map-based visualization.

Publication: H. Behrens, K. S. Candan, A. Gadkari, and others, "DataStorm-FE: A Data- and Decision-Flow and Coordination Engine for Coupled Simulation Ensembles", Very Large Databases (VLDB) 2018.
VLDB 2018 Publication Link